CT and MR of the Brain in Disorders of the Propionate and Methylmalonate Metabolism

Jan Brismar and Pinar T. Ozand

PURPOSE: To present the CT and MR findings in children with propionic and methylmalonic acidemia. METHODS: Twenty-three new patients with m ethylmalonic and 20 with propionic acidemia were examined with CT and/or MR of the brain. In total 52 CT and 55 MR studies were done. Twenty-six previously published cases were also reviewed. RESULTS: The findings were similar in the two syndromes. During the first month of life the examinations were either normal or showed white matter attenuation. Later during the first year moderate or even severe widening of sulci and fissures was seen, especially in infants with propionic acidemia. During therapy, these changes often resolved, especially in the patients with m ethylmalonic acidemia. Mild to m oderate delay in myelination was also a common finding in both disorders. Basal ganglia changes, predominately in the globus pallidus, were seen in five patients with methylmalonic acidemia and in two children with propionic acidemia; in two patients these changes were transient. CONCLU­ SION: Children who have methylmalonic or propionic acidemia, in addition to widening of cere­ brospinal fluid spaces and some delay in myelination, also often show symmetri c involvem ent of the basal ganglia.

Index terms: Acidemia; Brain, computed tomography; Brain, magnetic resonance; Brain, m etab­ olism; Pediatric neuroradiology

AJNR Am J f'feuroradio/15:1459-1473, Sep 1994

In inborn errors of metabolism, metabolic MMA and PPA are, like other organic aci­ routes are blocked, resulting in an accumulation demias, associated with cerebral complica­ of intermediary products that sometimes tions. Reports on the neuroradiologic findings causes severe neurologic symptoms. A major mainly consist of case reports: we found 13 metabolic pathway passes via propionyl-coen­ reported patients with MMA ( 2-8) and 13 with zyme A (propionyl-CoA) through methylmalo­ PPA (2, 9-12). nyl-CoA to the tricarboxylic cycle. Different We have had the opportunity to perform metabolic defects may block the breakdown of magnetic resonance (MR) and/or computed to­ propionyl-CoA, causing propionic acidemia mography (CT) of the brain in 20 new patients (PPA), or the breakdown of methylmalonyl­ with PPA and 23 with MMA. The purpose of the CoA, causing methylmalonic acidemia (MMA). present study was to combine our material with These two disorders are both among the least previously published cases, to study the time rare of the organic acidemias; by 1989 more course for development of neuroradiologic find­ than 100 cases of each disease had been ings, to study the distribution of lesions, and to reported ( 1). follow the response to therapy.

Received August 25, 1993; accepted after revision February 28, 1994. Subjects and Methods From the Departments of Diag nostic Radiology (J .B.) and Pediatrics and Biologica l and Medical Research (P.T.O.), King Faisal Specialist Hos­ We have, in 4 years, verified the diagnoses of PPA in 29 pital and Research Centre, Ri yadh, Saudi Arabia. patients and of MMA in 35. In all cases, the diagnoses were Address reprint requests to J an Brismar, MD, PhD, Department of based on both gas chromatography and mass spectros­ Radiology, King Faisal Specialist Hospital and Research Centre, PO Box copy and enzyme or isotope fixation studies. 3354, Ri ya dh 11 21, Saudi Arabia. All our cases of PPA were caused by propionyi-CoA­ AJNR 15:1459-1473, Sep 1994 0195-6108/94/1508-1459 carboxylase deficiency. All but two of our MMA cases © Ameri ca n Society of Neuroradiology responded dramatically to cobalamin and thus belong 1459 1460 BRISMAR AJNR: 15, September 1994 either to the cobalamin A or cobalamin B variants; the two findings improved in 7 of the children with PPA that did not respond were severe disease and probably and in 9 of the patients with MMA. Two patients belong to the group that lacks methylmalonyi-CoA mutase with PPA and 5 with MMA died during the ob­ (for explanation see below). servation period; in another 3 patients with PPA In 20 of the patients with PPA and in 23 of the children the condition worsened; in 6 patients with each with MMA we performed CT and/or MR of the brain on one or more occasions. Clinical data on these patients are disease the clinical findings were unchanged. summarized in Tables 1 and 2. The age at each examina­ The deaths were mainly in patients not compli­ tion is presented in Figures 1 and 2. In many of the pa­ ant to the therapy. tients, the initial study was done because of a family his­ The size of ventricles and sulci was normal tory, before any symptoms or signs had developed. during the first month of life in both diseases, CT was performed with 1 0-mm-thick contiguous axial but although in the infants with MMA the white sections. Contrast enhancement {Uitravist, 2 mL/ kg body matter also appeared normal, in most infants weight) was used only in a few cases. with PPA low-density white matter changes All MR examinations were done on a 1.5-T device using were seen. Later during the first year, promi­ dual-echo T2-weighted (2000/ 40, 80/ 2 [repetition time/ nence of ventricles and sulci was seen in all echo time/excitations]), T1-weighted (600/ 20/2) 7-mm­ patients, but white matter changes were less thick axial sections with 0- to 2.5-mm gap {depending on head size), and T1-weighted sagittal sections. In some of frequently diagnosed. Some delay in myelina­ the MR figures, the signal intensity is slightly unequal from tion was seen in most children with MMA exam­ side to side for technical reasons. ined during the second year of life; this was less Both the CT and the MR examinations were evaluated obvious in the infants with PPA. In children ex­ for prominence of ventricles, sulci, or fissures, and for the amined after 2 years, the findings were less anatomic extent of increased white matter T2 signal (or pronounced, and in several, previously delayed delayed myelination), both subjectively graded as absent, myelination or prominent ventricles and sulci slight, moderate, or severe. had normalized. As is evident from the tables, there seems to be a correlation between the degree of compliance to therapy and the sever­ ity of the radiologic findings. In some patients, Results despite therapy, a large number of hospital The results are summarized in Tables 1 and admissions, also to intensive care units, was 2. Although only 3 of the patients with PPA were needed for the treatment of the disease. Aston­ diagnosed after 9 months of age, the median ishingly, despite this fact, neuroradiologic find­ age for diagnosis in the MMA group was 12 ings were often normal or only mildly abnormal. months when those diagnosed as newborns, be­ Basal ganglia changes were seen in seven pa­ cause of family history, are excluded. In both tients and do not seem to correlate clearly either diseases, the children were often small, with to severity of clinical findings, to other radio­ small heads at diagnosis. The height was -2 SD logic abnormalities, to compliance to therapy, or less in 10 of 20 children with PPA, and in 14 or even to therapeutic problems as documented of 23 with MMA, head circumference was -2 by hospital admissions after the diagnosis. Al­ SD or less in 13 of the children with PPA and in though in MMA the basal ganglia lesions in all 10 of those with MMA. Clinical findings ranged but one patient were limited to globi pallidi, in from normal to severe hypotonia and choreo­ the two patients with PPA the changes involved athetosis in both diseases. At the diagnosis, also the caudate heads and the putamina. many of the children already had been admitted to the hospital multiple times, often also to in­ Discussion tensive care units. One boy with MMA had 11 admissions (3 of those to intensive care units) The incidence of MMA from one neonatal before diagnosis at 16 months of age; one girl screening program has been estimated to be 1 with PPA was admitted 6 times before diagnosis in 48 000 (13), but when less-severe cases are at 15 months. The compliance with therapy was also included, it may be as high as 1 in 25 000 variable. In only about half the patients was the (14). In the Massachusetts screening program compliance listed as good or excellent; in one only 1 patient with PPA was found among third it was poor or none. The finding of reduced 331 143 infants screened (15). In a paper de­ height or head circumference in most patients scribing 326 patients with inherited metabolic remained the same despite therapy; the clinical diseases, 21 cases of PPA compared with 31 of AJNR: 15, September 1994 METABOLIC DISORDERS 1461

Fig 1. Age at CT and MR examinations Propionic acidemia in 20 patients with PPA. Patient nr 20 -x X CT examination 19 -JO 18 JC) 0 MR examination 17 - 0 16 -x .... JO CT and MR examinations 15 --· JC) 14 x-x X 13 - 0 12 0 11 - 0 10 - X 0 9 - JC) 0 8 - X 7 0 ~X X 6 --~ 5 o-o 4 - - 0 X 3 - X - X 0 2 - X« 1 - X 1 2 3 4 Age in Years

Methylmalonic acidemia Patient nr 23 - X 22 - 21 -~-x• X CT examination 20 -o 19 x-o 18 - 0 • 0 MR examination 17 - 0 .. 16 - X) CT and MR examinations ~5 • 0 14 - 0 13 - 0 - 12 ~ 0 X 11 0 10 - 0 0 9 - 0 0 • 8 - 0 0 7 -o--o 6 10 0 0 5 - X: - 0 0 4 - X•o- 3 2 - .x 1 - • X -· 1 2 3 4 5 6 Age in Years

Fig 2. Age at CT and MR examinations in 23 patients with MMA...... TABLE 1: Clinical data and brain MR and/ or CT findings in 20 patients with PPA ,1::. (j) N Hospital CT and MR Findings Admissions Age at before/ after IJJ Sex/ Age Compliance Last Findings at Basal ;;o Clinical Findings Diagnosis Su lci, fissures, ventricles Subcortical white matter Patient at with Diet/ Visit Last Visit or ganglial Ui at Diagnosis (of Those, Diagnosis Medication (Age at at Death lesions Intensive ~;;o Death} 1-11 1-11 Care Unit < 1 mo 1 y 2: 2 y < 1 mo 1 y 2: 2 y mo mo Admissions) F/ 9 mo Height and HC - 4 ... 4(0}/ ... (9 mo) Died shortly after + - N None SD, severe diagnosis hypotonia, severe chorea, brain stem blind and deaf 2 M/7 mo Height or HC not ... / good, 2(1}/ 0 (12 mo) Height or HC not ++ -- N - None known, stupor, resp known, myoclonic seizures opisthotonus, blind, deaf, clonus 3 M/ 4 mo Height 75th percentile, Excellent 2(0}/ 2(0) 5 y 7 mo Height 50th, HC 1Oth ++->+ N N +->++ ++->+ N None HC - 2 SD, severe percentile, mild developmental developmental retardation retardation, deaf 4 M/ 8 mo Height - 4 SD, HC Poor/ 3(2}/ 4(2) 4y Height -3 SD, HC - ++ + + + + N None - 3 SD, severe moderate, - 4 SD, severe hypotonia resp hypotonia, chorea 5 F/ 3 mo Height and HC - 2 None 3(2)/ 2(2) (16 mo) Height and HC - 3 SD, - N ->+ - N - None SD, blind, unchanged clinica l nystagmus, severe findings hypotonia 6 F/ 1 mo Height 25th percentile, ... 2(2)/ ... (1.5mo) Died shortly after N - - +t+ - - - None HC - 3 SD, stupor, diagnosis hypotonic, coma 7 F/1 mo Height and HC - 2 Poor 1(1)/1(1) 3y Height and HC - 4 - +t+ ++ - N N - None SD, coma, severely SD , blind, deaf, no impaired milestone 8 M/1 mo Height - 3 SD, HC Fair 1 (0)/ 9(3) (15 mo) Height -6 SD , HC N +t+ +t+ N N N - None - 4 SD, intracranial )> - 2 SD, hypotonia c.... bleed led to :z hydrocephalus; ~ poliomyelitis, severe \.]1 hypotonia (j) 9 F/15 mo Height -4 SD, HC None/ fair, 6(0)/ 0 3 y 10 Height -3 SD, HC - - ++ ++ - - ++-> N Caudate, (1) -5 SD, severe resp mo - 2 SD, moderate globus "0..... (1) progressive hypotonia, pallidus, 3 pyramidal tract hemiplegia, putamen 0" ..,(1) disease, dysconjugate gaze ...... choreoathetosis 1.0 1.0 ,1::. TABLE 1: Continued )> c.....:z Hospital :::0 CT and MR Findings Admissions ...... Age at before/ after \Jl Sex/ Age Compliance Last Findings at Basal - Clinical Findings Diagnosis Sulci, fissures , ventricles Subcortical white matter (/) Patient at with Diet/ Visit Last Visit or ganglial !D at Diagnosis (of Those, \J Diagnosis Medication (Age at at Death lesions ..... Intensive !D Death) 1-11 1-11 Care Unit < 1 mo 1 y 2: 2 y < 1 mo 1 y 2: 2 y 3 mo mo 0'" Admissions) !D.., ...... 10 M/18 mo Height - 3 SD, HC - 2 Fair 4(1)/ 3(0) 3 y 8 mo Height - 4 SD, HC + + N + None \0 \0 SO, severe 50th percentile, hypotonia, severely moderate hypotonia """ delayed development 11 F/ 6 mo Height 75th and HC Excellent 1 (1 )/ 0 2.5 y Height 90th and HC - - + - - N Caudate, 1Oth percentile, 50th percentile, globus hypotonia, severe hypotonia, mild pallidus, choreoathetosis chorea putamen 12 F/ 4 mo Height and HC 40th Good 3(0)/ 2(0) 2y Height 25th and HC ++ N - N + None percentile, stupor, 5th percentile, mild hypotonia, clonus hypotonia 13 F/ 3 y Height and HC 50th None/ 1 (0)/ 0 5 y 6 mo Height and HC 90th - N - N None 8 m o percentile, demented, excellent, percentile, normal severe spastic resp quadriplegia 14 M/1 wk Height and HC - 2 Good 2(1 )/ 2(0) 22 mo Height and HC +++ -->++ - N None SO , slight hypotonia unchanged, almost normal 15 F/7 m o Height and HC - 3 Good 2(1 )/ 4(0) 26 mo Height and HC - 2 ++ ++ - + N None SO , hypotonia SD, same findings 16 F/ 4 m o Height - 3 SD, HC Good 2(1)/ 5(3) 9 mo Height - 5 SO, HC N - + - - None - 4 SO, hypotonia - 6 SO , severe hypotonia, severe choreoathetosis 17 F/ 1 mo Height 1Oth and HC Good 2 (1 )/ 1 (0) 15 mo Height and HC N + - ++ - + None 3 rn 5th percentile, unchanged, normal -; hemiplegia )> N - None OJ 18 M/ 2 m o Height 25th percentile, Poor 2 (1)/ 2(0) 9 mo Height 90th + - - 0 HC - 2 SD, blind, percentile, HC no r hypotonia, decrea sed change, unchanged n reflexes findings 0 N - - + - None Ui 19 M/ 1 wk Height and HC 40th Excellent 1 (1 )/ 0 4 mo Too early to as sess 0 percentile, cortical :::0 0 gras p, mild rn hypotonia :::0 (/) 20 M/ 5 d Head 75th and HC 2( 1)/ 0 5 wk Too early to assess N N - - None 50th percentile, mild hypotonia 0\""" Note.- HC indica tes head circumfe rence; N, normal findings; + , slight changes; ++, m oderate changes;+++, severe changes; -, no study during the age interva l. w TABLE 2: Clinical data and brain MR and/ or CT findings in 23 patients with MMA ..,. ..,.0"> Hospital CT and MR Findings Admissions Age at before/ after Sex/Age Compliance Last Findings at Basal co Clinical Findings Diagnosis Sulci, fissures, ventricles Subcortical white matter AJ Patient at with Diet/ Visit Last Visit or ganglial at Diagnosis (of Those, Ui Diagnosis Medication (Age at at Death lesions 3 Intensive )> Death) 1-11 1-11 AJ Care Unit < 1 mo 1 y ~ 2 y < 1 mo 1 y ~2 y mo mo Admissions)

M/1 0 mo Height - 4 SD, HC Excellent 2(2)!12(2) 7y Height - 5 SD, HC +t + N N N N Globus 50th percentile, no change, pallid us pyramidal signs hemiplegia 2 M/1 1 m o Height and HC None 1(1)/ 2(2) (12 mo) No change +t N - None unknown, hypotonia central hypoventilation, extrapyramidal signs 3 F/1 y Height - 3 SD, HC Excellent 3(1)/l (0) 5.5 y Height 50th +t--'>+ + - - N + None 90th percentile, mild percentile, HC no pyramidal signs change, normal 4 F/7 mo Height - 4 SD, HC Excellent 1(1)/ 2(0) 4.5 y Height - 5 SD, HC +--'>+t N N N +t + None - 4 SD, severe - 2 SD, mild hypotonia, poor chorea sucking 5 M/ 2 y Height - 3 SD, HC Good 5(5)/ 9(1) 6 y 10 Height - 5 SD , HC + +t Transient 10 mo - 2 SD, normal mo no change, normal basal neurology neurology ganglia 6 F/ 2 y Height - 3 SD, HC Fair 7(1)/16(2) 6y Height - 2 SD, HC + + None - 5 SD, hypotonia, - 3 SD , normal clonus 7 M/ birth Small with small head, Fair ... (1 y) Height - 4 SD, HC N + N N - None hypotonia 40th percentile, hypotonia 8 M/1 6 mo Height - 6 SD , HC Fair 11 (3)/ 4(0) 30 mo Height - 3 SD, HC - + N + + None - 2 SD, optic 25th percentile, atrophy, hypotonia, optic atrophy, choreoathetosis mild otherwise normal )> mental retardation c_ - N N N Globus :z 9 F/1 7 mo Height - 4 SD, HC Poor 1(1)/ 8(2) (3.5 y) Height and HC + AJ 50th percentile, unchanged, normal pallidus hypotonia, mild \Jl

mental retardation (fJ Unchanged normal - N - N None (!) 10 F/ 2 y Height 25th and HC Excellent 2(1 )/ 0 6y i:l,.... 10 mo 50th percentile, no (!) 3 neurology o- 11 M/ birth Height 1Oth and HC Excellent 1 (1 )/ 0 3.5 y Unchanged normal N + + N + + None (!).., 40th percentile, no ...... \0 neurology ..,.\0 TABLE 2: Continued )> c.... 7. Hospital CT and MR Findings ?? Admissions Age at ...... before/ after \.11 Sex/ Age Compliance Last Findings at Basal . Clinical Findings Diagnosis Sulci, fissures, ventricles Subcortical white matter (/) Patient at with Diet/ Visit Last Visit or ganglial at Diagnosis (of Those, II> Diagnosis Medication (Age at at Death lesions iJ..... Intensive II> Death) 1-11 1-11 Care Unit < 1 mo 1 y ~ 2 y < 1 mo 1 y ~ 2 y 3 mo mo 0" Admissions) ....II> ...... 12 M/ 9 d Height - 2 SD, HC Fair 1 ( 1)/ 0 3y Height - 3 SD, HC N + N N + N None lO 25th percentile, - 2 SD, normal lO severe "'" encephalopathy 13 M/1 y Height 50th percentile, None 2( 1 )/4(1) (2.5 y) Height and HC - - + + + ++ None HC - 3 SD, no unchanged, neurology hemiplegia 14 M/3 mo Height and HC 25th 1 ( 1 )/ ...... + - N Non,e percentile, hypotonia 15 F/ 3 d Height and HC -2 SD, Fair 1(1)/ 9(0) 3.5 y Height - 3 SD, HC + + None hypotonia, seizures no change, normal 16 F/3 mo Height - 3 SD, HC - 4 None 2(1 )/1 (1) (6 mo) No change +++ - - - N None SD, severe pyramidal signs, severe dysmorphia 17 F/ 12 mo Height - 5 SD, HC - 3 Excellent 3(2)/2(0) 34 mo Height - 4 SD, HC - N N - + + Globus SD, severe - 2 SD, mild pallid us hypotonia, hypotonia choreoathetosis 18 M/7 d Height and HC 25th Excellent 1 (1 )/0 2y No change - ++ - N None percentile, no neurology 19 M/14 mo Height - 6 SD, HC Excellent 5(4)/ 0 36 mo Height - 3 SD, HC - + N + N Globus 50th percentile, no change, mild pa llidus hypotonia, severe hypotonia 3: choreoathetosis [T1 -l 20 F/ birth Height and HC 50th ... 1 ( 1 )/...... N - - N - None )> OJ percentile, no 0 neurology r 21 F/ 5 d Height 25th and HC Poor 2(1 )/ 4(1) 18 mo Height - 4 SD, HC N + - N N -- None n 50th percentile, - 2 SD, mild 0 Ui hypotonia hypotonia 0 22 M/ 6 mo Height - 3 SD, HC - 2 Poor 3(2)/ 2(1) 20 mo Height and HC - 4 - ++ - N - None ;;u 0 SD, severe midline SD, hypotonia [T1 hypotonia, increased ;;u (/) deep tendon reflexes 23 F/ 5 d Height 25th and HC Excellent 2(1)/ 3( 1) (14 mo) No change - + N - - None 50th percentile, ...... hypotonia 01 \.11"'" Note.-HC indicates head circumference; N, normal findings; + ,slight changes;++ , moderate changes; +++, severe changes;-, no study during the age interval. 1466 BRISMAR AJNR: 15, September 1994

MMA were diagnosed ( 16), suggesting that the periodically, at which time the central ner­ incidence of PPA should be somewhat lower vous system findings of pyramidal tract signs than that of MMA, in the range of 1 in 35 000 to worsen, and seizures appear, frequently in as­ 1 in 70 000. sociation with acidosis and hyperammonemia. Consanguineous marriages are very com­ If not diagnosed and treated promptly, the con­ mon in Saudi Arabia, and many disorders inher­ dition eventually will progress into irretractible ited in an autosomal recessive mode, such as coma, and the patient will die (1, 17-20). PPA and MMA, are therefore present with much The propionate-methylmalonate metabolism higher prevalence than reported from western is summarized in Figure 3. The amino acids countries. The clinical features of PPA and MMA isoleucine, valine, and threonine, the odd-chain are somewhat similar. Both diseases are char­ fatty acids, and the side chain of cholesterol are, acterized by failure to feed, frequent vomiting, through different intermediate steps, metabo­ compensated to overt metabolic acidosis, and lized to propionyi-CoA. The further breakdown frequent infections. However, the acidosis is al­ of propionyi-CoA to methylmalonyi-CoA de­ most always very severe and resistant to treat­ pends on the enzyme propionyi-CoA-carboxy­ ment in MMA and milder and more easy to Iase, which for its action needs biotin as cofac­ manage in PPA. Although both diseases cause tor. PPA may be caused either by a lack of central hypotonia, this is usually very severe in propionyi-CoA-carboxylase or by biotin defi­ PPA and milder in MMA. PPA almost always ciency. In such metabolic blocks, alternative causes immunodeficiency and severe thrombo­ pathways are used, resulting in accumulation of cytopenia during times of metabolic crisis; such different metabolites. Biotin deficiency, in addi­ manifestations are much less apparent in MMA. tion to causing PPA, also blocks the function of Both diseases cause acute metabolic crises other carboxylases, causing multiple carboxy-

Isoleucine Valine Methionine Threonine Odd chain fatty acids Cholesterol

Propionic acid Methylmalonic acid

Lactate MeCbl t~ PC Homocystein Methionine Pyruvate Oxaloacetate --

ACC Acetyl-GoA Malonyl-GoA Biotin

Leucine

Methylcrotony/-CoA Methylglutaconyi-CoA - Cholesterol -MCC

Fig 3. Schematic presentation of the propionate and methylmalonate m etabolism. For details see "Discussion." Numbers 1 to 5 denote possible metabolic blocks: 1, propionyl-CoA carboxylase deficiency; 2, biotin deficiency; 3, methyl-malonyl-CoA mutase deficiency; 4, mutations late in adenocylcobalamine synthesis (cobalamin A or cobalamin B mutations); and 5, mutations early in the synthes is also affecting methylcobalamine. AJNR: 15, September 1994 METABOLIC DISORDERS 1467 lase deficiency. Biotin deficiency will not be fur­ may have acute metabolic crises during minor ther discussed. MMA is caused by a block of the infections and after not eating. metabolism of methylmalonate-CcA to succi­ The neuroradiologic findings were similar in nyl-CoA either from lack or deficiency of the our patients with PPA and MMA but were usually enzyme methylmalonyl-CoA mutase or from a more severe in those with PPA. This was obvi­ Jack of the necessary coenzyme adenocylcobal­ ous even in the neonatal period, with all but one amine. Deficiency of the mutase causes a of the infants with MMA being relatively asymp­ milder form of disease than a total Jack. Lack of tomatic and identified mainly because of their adenocylcobalamine may be caused by meta­ family history. Two of the five infants with MMA bolic defects at different stages of its synthesis. examined during the first month of life were Defects at a late stage (cobalamin A and cobal­ clinically healthy, two had slight muscular hy­ amin B defects) cause only MMA; defects at an potonia, and only one had severe encephalop­ early stage, which also cause homocystinuria, athy. Although all five of these patients had are beyond the scope of this article. normal CT or MR findings, only two of six infants The age at presentation in patients with MMA with PPA examined with CT or MR of the brain differs depending on the underlying defect ( 1, during the same period of life had normal­ 19, 21). Whereas 80% of the children with a appearing white matter. Two had mild and one total Jack of mutase become ill during the first had moderate white matter abnormalities; the week and 90% during the first month of life, the sixth displayed severe changes with necrosis of majority of patients with partial methylmalonyl­ subcortical white matter (case 6; Fig 4). All six CoA mutase deficiency or with adenocylcobal­ had normal-size ventricles and sulci. amine defects present after the first month (20). Later during the first year of life, the sulci and fissures were widened in all 13 infants with MMA Except 7 patients, all 35 cases of MMA seen at and in all 11 with PPA that were studied: mark­ our hospital presented with symptoms after 1 edly in 1 with MMA and in 3 with PPA, moder­ month of age. Typically, patients with PPA ately in 7 with MMA and in 8 with PPA. Three of present within the first week of life. However, in the 13 children with MMA had initial normal some patients the disease manifests itself in late findings. In one child with PPA, shunt-requiring infancy and childhood, often in association with hydrocephalus developed secondary to intra­ a metabolic stress such as respiratory infections cranial bleeding caused by thrombocytopenia. or gastroenteritis. There are individuals who All but 3 infants, who had PPA, had normal­ have no clinical evidence but nonetheless are appearing white matter. Whereas all 4 infants diagnosed as having a total lack of propionyl­ with MMA in this group who were diagnosed and CoA-carboxylase because of family history. ( 1, treated with diet and medication since the neo­ 11, 17). The reason behind this variety of pre­ natal period had only slight prominence of the sentations for the same enzyme defect is not cerebrospinal fluid (CSF) spaces, 6 of 7 diag­ understood. Also, some children with MMA nosed later had moderate or severe ( 1 patient) caused by partial methylmalonyl-CoA mutase changes. Again, findings were more severe in deficiency may develop completely normally the infants with PPA: all 3 children diagnosed and remain asymptomatic (22). and treated from the first month of life, who Essentially, the treatment for both diseases were studied during this age interval, had severe consists of restriction of proteins while main­ widening of the CSF spaces despite the early taining adequate general nutrition. Patients with start of therapy. MMA with adenocylcobalamine defects respond Mild delay in myelination is difficult to identify dramatically to cobalamin. The treatment for during the first year of life . A delay in myelina­ MMA consists of a formula restricted in L-isoleu­ tion was verified in 9 of 10 children with MMA cine:L-carnitine (100-200 mg/kg per day), dur­ studied during the second year of life , mild in 8 ing the acute phase daily intramuscular injec­ and moderate in one. At a repeat study after 2 tions of 1 mg of hydroxycobalamine, and in the years of age myelination had improved in 3 of chronic phase 0.5 mg of cyanocobalamin daily 7 of these children. Also, the widening of the as nasal spray. For PPA the therapy is a formula extracerebral CSF spaces subsided with age. restricted in L-isoleucine and L-valine:biotin ( 10 Eight of 11 children with MMA studied between mg/kg per day) and L-carnitine ( 100-200 2 and 6 years of age had normal sulci and mg/kg per day). Even with treatment, patients fissures; in 6 children previously widened sulci 1468 BRISMAR AJNR: 15, September 1994

Fig 4 . One-m onth-old girl with PPA (case 6 ): microcephaly, muscular hypotonia, stupor, and com a. A , CT shows marked decrease in attenuation of subcortical white m atter. B and C, MR (T1-weighted) shows signs of white m atter necrosis with bilaterally small frontal hem orrhages. had returned to normal. In only 1 child an im­ years of age demonstrated moderate subcorti­ pairment, with progression of the white matter cal white matter disease sparing the most pe­ abnormalities, was observed. This child later ripheral white matter; at 4 years the putamina died in a neglected MMA coma; his parents were were bilaterally swollen and of irregularly, noncompliant with therapy. Five patients with markedly increased T2 intensity. A repeat MR 3 PPA were studied after the second year of life weeks later showed normalization of the basal (Fig 5): in 3 patients, previously delayed myeli­ ganglia changes and some improvement of the nation had normalized; in one patient mild T2 subcortical white matter disease. In the remain­ white matter changes had developed; and in the ing 4 children, symmetrical basal ganglia ne­ fifth patient the initial MR at almost 4 years of crosis with volume loss was demonstrated; in all age was normal despite severe clinical symp­ only the globi pallidi were involved (case 19, Fig toms. Two of these 5 children had normal ven­ 7). In 2 of the patients the necrosis was present tricles and sulci, 3 slightly or moderately di­ at the first examination (cases 9 and 19, at 8 lated. and 17 months of age, respectively). One child One reason for the difference in severity of (case 17) had a normal MR at 17 months and neuroradiologic findings between our patients showed basal ganglia necrosis 1 year later; with PPA and MMA may be that all but 2 of our one (case 1) had several normal CT studies children with MMA had mild variants of the up to 2.5 years and at 4 years had developed disease (probably cobalamin A or cobalamin B necrosis. defects) that respond well to cobalamin; they Basal ganglia lesions were seen in two of our therefore initially had fewer and less-severe patients with PPA. In one infant, an MR study at metabolic crises than the patients with PPA. 1.5 years of age demonstrated increased T2 Despite this fact, basal ganglia lesions were intensity within the globus pallidus, the puta­ more common in the MMA group. No obvious men, and the caudate nucleus bilaterally (case correlation was found between the severity of 11 , Fig 8). In the second child the first CT and the disease and the occurrence of basal ganglia MR examinations at 16 months (case 9, Fig 5) changes. Although seen in only 1 of 12 patients showed low-density, high-T2 intensity swollen with MMA examined between 1 month and 1 globi pallidi, and marked prominence of sulci year of age, basal ganglia lesions were present and fissures and some delay in myelination. Ten in 5 of 10 children older than 28 months. In 1 months later the changes in the globi pallidi child the lesions involved the putamina and were less prominent, but increased T2 intensity were transient (case 5, Fig 6); CT and MR at 3 was now also present in the putamina and cau- AJNR: 15, September 1994 METABOLIC DISORDERS 1469

Fig 5. Girl with PPA (case 9 ), di ­ agnosed at 15 m onths of age, with severe progressive pyramidal tract disease and choreoathetosis. A, CT at 16 m onths shows m od ­ erate prominence of CSF spaces, mild increase in gray and white matter differentiation, and m arked swelling and decreased attenuation of the globi pallidi. B and C, MR (T1 - and T 2- weighted, respectively) 6 days later again shows lesions limited to the globi pallidi. 0 and £, Re peat MR (T1- and T2- weighted, res pectively) after 10 m onths shows the left les ion de­ creased in size , whereas on the ri ght side the entire lentiform nucleus as G H well as the caudate is now involved . F, CT after another 1 0 m onths fails to demonstrate any basal ganglia lesion, and also the widening of the CSF spaces is less prom inent. G and H, MR (T1- and T2-weighted , respectively) 10 m onths later shows sm all high T2 residues within the right lentiform nucleus and caudate head; the right lentiform nucleus also has lost volume. 1470 BRISMAR AJNR: 15, September 1994

A Fig 6. Boy with MMA (case 5): normal neurologic findings during the entire fol­ low-up period. A, At 3 years of age, MR (T2-weighted) shows moderate white matter disease but no basal ganglia changes. 8 and C, CT 4 years after a metabolic crisis shows markedly decreased density within the putamina but no involvement of other basal ganglia regions. D, At MR (T2-weighted) 8 days later, the putamina are less swollen and the le­ sions are clearly seen. £, MR (T2-weighted) after another 3 weeks no longer shows any basal ganglia disease; also, the subcortical white matter changes have decreased.

D E date heads bilaterally. At a CT study another 10 peared normal, but features of cerebral atrophy months later no basal ganglia disease was iden­ were present. Only two of the patients with MMA tified, but a repeat MR after 10 months again we found in the literature were examined before showed a slitlike lesion in the right globus pal­ 1 year of age. In one infant examined at 1 month lidus (Fig 5). This illustrates the difference in of age marked white matter edema was present sensitivity between the two imaging modalities; (2); in 1 asymptomatic sibling to a patient with the lesion was probably present all the time. MMA studied at 9 months, CT findings were We found CT and MR results reported from 13 normal (7). As in our material, the changes patients with PPA and 13 with MMA (2-10, 17, were less obvious in older children: mild to 18); MR was performed in 1 child with PPA ( 17) moderate atrophy was reported in 4, and mild and 2 children with MMA (7, 8). The findings white matter changes in 2 of 5 children with have been similar to those in our series. A report MMA studied during the second year of life on the CT findings in a patient with PPA ap­ (2-4, 6), but only 2 of 7 children studied with peared in 1981 (9): aCT scan at 2 weeks of age CT of the brain later during childhood showed showed diffusely decreased attenuation of the mild atrophy, and none showed white matter white matter; at 8 months white matter ap- changes (4-8). AJNR: 15, September 1994 METABOLIC DISORDERS 1471

A B c Fig 7. Boy with MMA (case 19): muscular hypotonia and severe choreoathetosis. A, CT at 17 months of age shows slight prominence of CSF spaces and slightly decreased attenuation within the globus pallidus, bilaterally. Band C, MR (T1- and T2-weighted, respectively) beautifully demonstrates very well demarcated necrotic lesions with loss of volume in the globi pallidi and also shows mild prominence of CSF spaces. Other sections documented slight white matter disease.

Autopsy findings in two infants with PPA who of the posterior limb of the internal capsule, died during the neonatal period have been re­ globus pallidus, the ascending and descending ported (10, 23, 24). In a 12-day-old infant with fibers of the brain stem, the tegmentum, and the a fatal cerebellar hemorrhage, also seen at CT, posterior columns of the spinal cord (23). Vac­ patchy subcortical demyelination of subcortical uolization of the medial lemniscus, superior cer­ white matter, status spongiosus of the globus ebellar peduncles, and posterior columns were pallidus and internal capsule, and marked loss found to be especially severe. White matter of deep cerebellar subcortical myelinated fibers changes of this type would explain the MR find­ with spongy necrosis was found (10, 24). In a ings in our PPA case 6 (Fig 1 ), who died at 1.5 26-day-old neonate, who had died from meta­ months of age. bolic acidosis, microscopic examination of the All 11 published cases of MMA studied after brain demonstrated diffusely vacuolated myelin the first year of life had lesions affecting the

Fig 8 . Eighteen-month-old girl with PPA (case 11 ) : muscular hypotonia and severe choreoathetosis. A and B, MR (T2-weighted) shows slight prominence of sulci and increased intensity within the lentiform nucleus and caudate head bilaterally.

A B 1472 BRISMAR AJNR: 15, September 1994

globus pallidus. This need not mirror the true 28), but there is no definite explanation. Fur­ frequency of basal ganglia involvement; many thermore, the causes of the apparent difference of the cases were probably published because between MMA and PPA in the distribution of the of the unusual lesions in the globus pallidus. In basal ganglia changes are unknown. It may be our hospital an MR study of the brain, and often caused by specific toxic effects from different also a CT examination, is part of the routine metabolites or depend on the fact that different exam for possible neurometabolic diseases basal ganglia structures might be the metabol­ even in patients with minimal or no signs of ically most active structures in the brain in MMA brain involvement. The frequency of basal gan­ and PPA crises, and therefore are most vulner­ glia changes in our material therefore should able to anoxia and metabolic insults. better indicate the true incidence. In a review In the retrospective series on the outcome of article on MMA in 1987 the authors stated that PPA (18), 10 of 11 infants in the early-onset "several cases, as yet not fully reported in the group, with the most severe disease, were of literature, have been described to the authors of Arab or Indio-Pakistani origin; 8 of 9 in the infarcts in various parts of the brain at several late-onset group, with milder disease, were years of age" ( 14 ). In all published reports the European. All our patients are of Saudi Arabian basal ganglia changes, however, have been lim­ origin, but nonetheless the age at onset and the ited to globus pallidus and the adjacent parts of severity of disease varies markedly. This is in the internal capsules. agreement with other reports. A study on a One report described cerebral atrophy at CT massively inbred Mennonite-Amish community in a 3-year-old girl with PPA known since 2 with several patients with PPA, in which the months of age (2). A retrospective study on the disease in all probability originated from the neurologic outcome in PPA identified a total of same ancestor and thus probably was biochem­ 20 patients diagnosed during a 25-year period ically identical, showed a wide variety of clinical (18). They found the patients were divided into manifestations (21 ), ranging from severe dis­ two distinct groups: those with disease onset ease to complete lack of symptoms. · during the first week of life, and those with onset White matter abnormalities and prominent after the first 6 weeks. The disease was more CSF spaces are nonspecific findings, present in severe and the prognosis much worse in the many neurometabolic disorders. Basal ganglia early-onset group. CT was performed in 4 chil­ lesions are less frequently seen but may occur dren in each group-the patients' exact ages at not only in different organic acidemias but also the CT examinations were not given in the re­ in anoxic lesions, after methanol intoxication, port. In the early-onset group, CT in 3 cases and also in infectious diseases such as sym­ showed cerebral atrophy and in 1 was normal. metrical basal ganglia changes in herpes sim­ In 3 of the 4 patients with late-onset PPA, CT plex encephalitis. The finding of basal ganglia showed basal ganglia lesions; in 1 the study lesions, be they slitlike high-T2 intensity lenti­ was normal. In 1 child the lesion involved the form nuclei, shrunken caudate heads, or lentiform nuclei; in 1 also the caudate nuclei acutely swollen basal ganglia, always should were affected, and in the third the medial pos­ prompt examination for possible · underlying terior lentiform nucleus, the anterior half of the metabolic disease, so that therapy can be thalamus, and subcortical white matter were started and thereby, it is hoped, further dam­ bilaterally involved. In all patients the basal age to the brain avoided. If the lesions are lim­ ganglia lesions resolved during 1 to 3 months; ited to the globi pallidi, MMA is the likely in 1 child further basal ganglia lesions later cause, but the diagnosis nevertheless has to developed. Microscopic examination of the be confirmed biochemically. white matter in a 23-month-old infant who died in a metabolic coma secondary to PPA (24) revealed occasional regions of perivascu­ References lar spongy rarefaction in the caudate nucleus 1. Rosenberg LE, Fenton WA. Disorders of propionate and methyl· but otherwise well-preserved myelin without malonate metabolism. In : Scriver CR, Beaudet AL, Sly WS, Valle spongiform changes. D, eds. Th e Metabolic Basis of Inherited Disease. 6th ed. New York: McGraw-Hill, 1989:821-842 The reason that the globus pallidus is the 2. Gebarski SS, Gabrielsen TO, Knake JE, Latack JT. Cerebral CT most vulnerable ganglia structure in MMA has findings in methylmalonic and propionic acidemias. AJNR Am J been discussed in several previous reports (25- Neuroradio/1983;4:955-957 AJNR: 15, September 1994 METABOLIC DISORDERS 1473

3. Korf B, Wallman JK, Levy HL. Bilateral lucency of the globus 17. Shigematsu Y, Mori I, Nakai A, et al. Acute infantile hemiplegia in pallidus complicating methylmalonic acidemia. Ann Neuro/1986; a patient with propionic acidemia. Eur J Pedialr 1990; 149:659- 20:364-366 660 4. Heidenreich R, Natowicz M, Hainline BE, et al.- Acute extrapyra­ 18. Surtees RAH, Matthews EE, Leonard JV. Neurologic outcome of midal syndrome in methylmalonic acidemia: "metabolic stroke" propionic acidemia. Pediatr Neurol 1992;8:333-337 involving the globus pallidus. J Pediatr 1988; 1 13:1 022-1027 19. Wolf B, Hsia YE, Sweetman L, Gravel R, Harris DJ, Nyhan WL. 5. Thompson GN, Christodoulou J, Danks DM. Metabolic stroke in Propionic acidemia: a clinical update. J Pedialr 1981 ;99:835- 846 methylmalonic acidemia (letter). J Pediatr 1989; 115:499-500 20. Matsui SM, Mahoney MJ, Rosenberg LE. The natural history of the 6. de Sousa C, Piesowicz AT, Brett EM, Leonard JV. Focal changes inherited methylmalonic acidemia s. N Eng/ J Med 1983;308:857- in the globi pallidi associated with neurological dysfunction in 861 methylmalonic acidemia. Neuropediatrics 1989;20: 199-201 21. Kidd JR, Wolf B, Hsia YE, Kidd KK. Genetics of propionic aci­ 7. Roodhooft AM, Baumgartner ER, Martin JJ, Blom W, van Acker demia in a Mennonite-Am ish kindred. Am J Hum Gen et 1980;32: KJ. Symmetrical necrosis of the basal ganglia in methylmalonic 236-245 acidemia. EurJ Pediatr 1990;149:582-584 22. Ledley FD, Levy HL, Shih VE, Benjamin R, Mahoney MJ. Benign 8. Andreula CF, De Blasi R, Carella A. CT and MR studies of meth­ methylmalonic aciduria. N Eng/ J Med 1984;311: 1015-1018 ylmalonic acidemia. AJNR Am J Neuroradio/1991;12:410-412 23. Shuman RM , Leech RW, Scott CR. The neuropathology of the 9. Barnes DM, Enzmann DR. The evolution of white matter disease nonketotic and ketotic hyperglycinemias: three cases. Neurology as seen on computed tomography. Radiology 1981; 138:379-383 1978;28: 139-146 10. Dave P, Curless RG , Steinman L. Cerebellar hemorrhage compli­ 24. Steinman L, Clancy RR , Cann H, Urich H. The neuropathology of cating methylmalonic and propionic acidemia. Arch Neuro/1984; propionic acidemia. Dev Med Child Neuro/1983;25:87- 94 41:1293-1296 25. Rosenberg NL. Methylmalonic acid, methanol, metabolic acido­ 11. Ozand PT, Gascon GG . Organic acidurias: a review. Part 1. J sis, and lesions of the basal ganglia. Ann Neurol 1987 ;22:96 Child Neuro/1991;6:196-219 26. Korf BR, Wallman JK, Levy HL. Methylmalonic acid, methanol, 12. Ozand PT, Gascon GG. Organic acidurias: a review. Part 2. J metabolic acidosis, and lesions of the basal ganglia (reply to Child Neuro/1991 ;6:288-303 letter) . Ann Neuro/ 1987;22:96-97 13. Coulombe JT, Shih VE, Levy HL. Massachusetts metabolic disor­ 27. Berry GT, Heidenreich R. Metabolic stroke in methylmalonic aci­ ders screening program, II: methylmalonic aciduria. Pediatrics demia (reply to letter). J Pediatr 1989;115:500 1981;67:26-31 28. Naidu S, Moser HW. Value of neuroimaging in metabolic diseases 14. Mahoney MJ, Bick D. Recent advances in the inherited methyl­ affecting the central nervous system. AJNR Am J Neuroradiol malonic acidemias. Acta Pediatr Scand 1987;76:689-696 1991 ;12:413-416 15. Levy HL. Newborn screening for metabolic disorders. N Eng/ J Med 1973;288:1299-1300 16. Saudubray JM, Ogier H, Bonnefont JP, et al. Clinical approach to inherited metabolic diseases in the neonatal period: a 20-year survey. J Inherited Metab Dis 1989; 12 (Suppl 1 ):25-41